Community Memory: 1974
The Community Memory project was an experiment in providing the public a free-to-use computer system to read and post messages, on any subject, anonymously. The first terminal was placed in a music store in Berkeley, California in 1973. During the first incarnation of the project (August 1973-January 1975), terminals were placed in three San Francisco locations and two Vancouver locations.
This website has translated the text from the scanned printouts collected by the Computer History Museum and donated by Lee Felsenstein into a browseable, linked form. The printouts are a snapshot of posts made to the system in February and March 1974 at the Leopold's Record Store terminal.
Discover how Berkeley residents in 1974 interacted with a public, digital bulletin board, without any limitations on their speech or use of the system.
Posts are organized within a top-level "Directory" and by keyword.
| Body | Change | Children | CM |
| Community | Connect | Electronics | Energy |
| Food | Head | Letter | Media |
| Music | Place | Play | Sale |
| Spirit | Voice | Want | Women |
Frequently Asked Questions (FAQ)
How did the public interact with the Community Memory hardware?
The project used donated hardware and made modifications to make the terminals more visible and robust to public use. One of the designers describes the terminal at Leopold's Records as:
The terminal was one of our donated Teletype 33s, a device too noisy for use in public spaces. We had already dealt with the silencing issue at our office, and I hand-built a cardboard enclosure insulated inside with urethane foam and a Velcro-secured clear flexible vinyl flap as a window.
Two holes on the front of the box with star-cutout vinyl gaskets allowed a user to reach through to the keyboard. A small vent fan prevented overheating and we hand-lettered "Community Memory" on the side in psychedelic script. This sat on a card table with a chair in front and another to the side where our minder sat.
Me and My Big Ideas by Lee Felsenstein, pg 80-81
A Hazeltine video terminal was installed early in 1974, although a CM post called it "flakey".
Where could the public access Community Memory?
The pilot location was Leopold's Records, a non-profit community record and music store in Berkeley, CA frequented by students and musicians. Later terminals were placed in a public in San Francisco's Mission District and a Whole Earth access store. The system was also tested in Vancouver, British Columbia. Terminals were placed into a public library and a community information center.
Where can I learn more about Community Memory?
Online sources include:
- Community Memory: Precedents in Social Media Movements
- An introduction to the Community Memory project and artifacts available at the Computer History Museum
- Community Memory: A Public Information Network
- Contemporary (1976) description of the system and how it is being used. Alternate Creative Computing version
- Implications of community memory
- Contemporary (1976) description of the philosophy and design for a community information system
- The Community Memory Project: Description and Current Status
- A 1978 article providing a retrospective on the first incarnation of Community Memory and plans for the future
Books covering the project include Steven Levy's Hackers: Heroes of the Computer Revolution, John Markoff's What the Dormouse Said: How the Sixties Counterculture Shaped the Personal Computer Industry, and Lee Felsenstein's Me and My Big Ideas: Counterculture, Social Media, and the Future.
How was the data on this site assembled?
We started with two scanned PDFs preserved by the Computer History Museum: artifact 102703230-05-0002-acc and 102703230-05-0003-acc (catalog link). These print-outs represent posts made at the Leopold's Records terminal in Berkeley, CA, in February and March 1974. We eliminated pages from the two PDFs that did not contain printouts of user posts, which as a side effect removed the existing OCR text. (The OCR text in the original PDFs contains many errors.) Artifact 102703230-05-0005-acc also contains user posts, but they are a duplicate of 102703230-05-0003-acc.
Using Google's Document AI service (processor pretrained-ocr-v2.1-2024-08-07), we extracted text from each of the PDFs. We experimented with using a language hint, but found it did not improve the output.
We further cleaned the textual output by converting it to an ASCII encoding constrained to glyphs used by the SDS-940 computer. For example, the OCR process converted some 0 glyphs as Greek Theta and X glyphs as Greek Chi.
Finally, we manually went through the files, fixing structural issues within the files (e.g. starting points of posts, directory declarations, and keyword lists), as well as correcting whitespace within some posts to restore original intent. A script parsed the final corrected text and generated a JSON file used by the website's template system to generate the final HTML representation.
How can I contact the owners of this site?
Email admin@communitymemory1974.forum